03¹Ú³ë¿í7~272s

Size: px

Start display at page:

Download "03¹Ú³ë¿í7~272s"

다민 주
5 years ago
Views:

1 Korean Journal of Remote Sensing, Vol.23, No.4, 2007, pp.257~272 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes No-Wook Park*, Hoonyol Lee**, and Kwang-Hoon Chi* *Geoscience Information Center, Korea Institute of Geoscience and Mineral Resources **Department of Geophysics, Kangwon National University Abstract : This paper analyzed relationships between various features from SAR data with multiple acquisition dates and mode (frequency, polarization and incidence angles), and land-cover classes. Two typical types of features were extracted by considering acquisition conditions of currently available SAR data. First, coherence, temporal variability and principal component transform-based features were extracted from multi-temporal and single mode SAR data. C-band ERS-1/2, ENVISAT ASAR and Radarsat-1, and L-band JERS-1 SAR data were used for those features and different characteristics of different SAR sensor data were discussed in terms of land-cover discrimination capability. Overall, tandem coherence showed the best discrimination capability among various features. Long-term coherence from C-band SAR data provided a useful information on the discrimination of urban areas from other classes. Paddy fields showed the highest temporal variability values in all SAR sensor data. Features from principal component transform contained particular information relevant to specific landcover class. As features for multiple mode SAR data acquired at similar dates, polarization ratio and multi-channel variability were also considered. VH/VV polarization ratio was a useful feature for the discrimination of forest and dry fields in which the distributions of coherence and temporal variability were significantly overlapped. It would be expected that the case study results could be useful information on improvement of classification accuracy in land-cover classification with SAR data, provided that the main findings of this paper would be confirmed by extensive case studies based on multi-temporal SAR data with various modes and ground-based SAR experiments. Key Words : Land-cover classification, Polarization, Coherence, Temporal variability. hoonyol@kangwon.ac.kr 257

2 Korean Journal of Remote Sensing, Vol.23, No.4, 2007 ä 258

3 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes ü 259

4 Korean Journal of Remote Sensing, Vol.23, No.4, 2007 Table 1. Multiple SAR sensor data used for illustrating main characteristics of features from single polarization/multi-temporal SAR data. Interferometric pairs and perpendicular baseline distance for coherence are listed in parentheses. Sensor Sensor JERS-1 SAR Wavelength/IncidenceAcquisition Acquisition dateextracted features angle/polarization (year/month/day) (year/month/day) /11/ /01/ /02/26 L band (23.5cm)/35 /HH /04/ /05/ /07/ /08/ /10/ /12/21 ERS-1/2 Tandem C band (5.7cm)/23 /VV /12/ /01/ /02/ /03/ /04/24 ENVISAT ASAR C band (5.7cm)/23 /VV /05/ /07/ /08/ /09/ /10/ /04/ /04/ /05/ /06/12 Radarsat-1 C band (5.7cm)/40 /HH /07/ /07/ /08/ /09/ /10/10 Extracted features coherence (12: 969m, 13: 973m, 14: 403m, 23: 1931m, 24: 1372m, 34: 587m, 78: 566m), temporal variability, principal components Tandem coherence (12: 248m) coherence (12: 696m, 23: 538m, 46: 275m, 78: 186m, 79: 115m, 89: 71m), temporal variability, principal components coherence (12: 848m, 24: 130m, 35: 142m, 46: 73m, 78: 739m, 89: 98m), temporal variability, principal components 260

Coherence images and their histogram distributions with

pair, (b) Radarsat-1, (c) ENVISAT ASAR, and (d) JERS-1.

5 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes (a) (b) (c) Fig. 1. Coherence images and their histogram distributions with respect to five land-cover classes, (a) ERS-1/2 Tandem pair, (b) Radarsat-1, (c) ENVISAT ASAR, and (d) JERS-1. Typical land-cover areas are expressed with arrows and captions. 261

6 Korean Journal of Remote Sensing, Vol.23, No.4, 2007 Fig. 1. Continue (d) 262

histogram distributions with respect to five

7 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes (a) (b) (c) Fig. 2. Temporal variability images and their histogram distributions with respect to five land-cover classes, (a) JERS-1, (b) Radarsat- 1, and (c) ENVISAT ASAR. 263

8 Korean Journal of Remote Sensing, Vol.23, No.4,

9 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes (a) (b) (c) Fig. 3. Temporal variability values for various multi-temporal data combinations. (a) combination of April, May and July data, (b) combination of April, May, June and July data, and (c) combination of June, August, September and October data. Fig. 4. Qualitative relationships between temporal variability and coherence for land-cover classes. 265

10 Korean Journal of Remote Sensing, Vol.23, No.4,

11 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes (a) Fig. 6. The second principal component values of multiple coherence images extracted from multitemporal JERS- 1, Radarsat-1 and ENVISAT ASAR data sets, and their histogram distributions with respect to five land-cover classes. (b) Fig. 5. (a) Principal component loadings of the first three principal components from 9 Radarsat-1 data sets (from Park and Chi, 2007), (b) their histogram distributions with respect to five land-cover classes. 267

12 Korean Journal of Remote Sensing, Vol.23, No.4, 2007 ü ü 268

classes, (b) HH/VV polarization ratio and histogram distributions with respect to five land-cover

13 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes (a) Fig. 7. (a) VH/VV polarization ratio and histogram distributions with respect to five land-cover classes, (b) HH/VV polarization ratio and histogram distributions with respect to five land-cover classes. (b) N S i i N S i w k m ik s ik w k S i m ik s ik S i S _ i m ik S i = (1) s ik w k S i N S i 269

Korean Journal of Remote Sensing, Vol.23, No.4, 2007 Table 2. Multi-mode SAR data used for extracting dualpolarization ratio and multiple relative variability.

/F2/HH 2005/06/12 C/34 /S3/HH 2005/06/19 (a) (b) Fig. 8.

14 Korean Journal of Remote Sensing, Vol.23, No.4, 2007 Table 2. Multi-mode SAR data used for extracting dualpolarization ratio and multiple relative variability. Sensor Band/Incidence angle/ Acquisition date Mode/Polarization (year/month/day) ENVISAT C/23 /IS2/HH&VV 2005/05/29 ASAR C/19 /IS1/VH&VV 2005/06/17 Radarsat-1 C/40 /F2/HH 2005/06/12 C/34 /S3/HH 2005/06/19 (a) (b) Fig. 8. (a) multi-channel relative variability and histogram distributions with respect to five land-cover classes, (b) the second principal component computed from six relative backscattering coefficient images and histogram distributions with respect to five landcover classes. 270

15 Analysis of Relationships between Features Extracted from SAR Data and Land-cover Classes 271

16 Korean Journal of Remote Sensing, Vol.23, No.4, 2007 Bruzzone, L., M. Marconcini, U. Wegmüller, and A. Wiesmann, An advanced system for the automatic classification of multitemporal SAR images, IEEE Transactions on Geoscience and Remote Sensing, 42(6): Dammert, P. B. G., J. Askne, and S. K hlmann, Unsupervised segmentation of multitemporal interferometric SAR images, IEEE Transactions on Geoscience and Remote Sensing, 37(5): Engdahl, M. E. and J. M. Hyypä, Land-cover classification using multitemporal ERS-1/2 InSAR data, IEEE Transactions on Geoscience and Remote Sensing, 41(7): Gens, R. and J. L. van Genderen, SAR interferometry - issues, techniques, applications, International Journal of Remote Sensing, 17: Park, N.-W. and K.-H. Chi, Integration of multi-temporal/polarization C-band SAR data sets for land-cover classification, International Journal of Remote Sensing, accepted. Quegan, S., T. L. Toan, J. J. Yu, F. Ribbes, and N. Floury, Multitemporal ERS SAR analysis applied to forest mapping, IEEE Transactions on Geoscience and Remote Sensing, 38(2): Strozzi, T. and U. Wegmüller, Delimitation of urban areas with SAR interfermetry, Proc. of IGARSS 1998, Seattle, WA, pp Strozzi, T., P. B. G. Dammert, U. Wegmüller, J. M. Martinez, A. Beaudoin, J. Askne, and M. Hallikainen, Landuse mapping with ERS SAR interferometry, IEEE Transactions on Geoscience and Remote Sensing, 38(2): Ulaby, F. T. and M. C. Dobson, Handbook of Radar Scattering Statistics for Terrain, Artech House. Norwood, MA. Wegmüller, U., T. Strozzi, A. Wiesmann, and C. Werner, ENVISAT ASAR for land cover information, Proc. of IGARSS 2003, Toulouse, France, DVD publication. Wiesmann, A., U. Wegmüller, C. Werner, and T. Strozzi, ASAR multi-swath techniques, Proc. of IGARSS 2003, Toulouse, France, DVD publication. 272

09È«¼®¿µ5~152s

09È«¼®¿µ5~152s Korean Journal of Remote Sensing, Vol.23, No.2, 2007, pp.45~52 Measurement of Backscattering Coefficients of Rice Canopy Using a Ground Polarimetric Scatterometer System Suk-Young Hong*, Jin-Young Hong**,

03¹Ú³ë¿í7~272s

03¹Ú³ë¿í7~272s